Printer Friendly

Unification of China's foreign exchange rates.


On January 1, 1994, China took a major step toward currency convertibility by unifying its official and swap market exchange rates. A national foreign exchange market centered around banks replaced the system of trading foreign exchange through the foreign exchange adjustment centers. The major motivation behind the reform is to overhaul the fragmented swap market and to improve the efficiency of foreign exchange allocation. However, no one has published a study on the degree of market segmentation and the efficiency of the FEACs. Such a study would put the significance of the unification in perspective. In an attempt to provide this perspective, this paper evaluates the swap market's performance before the unification, examines the implications of the unification for the Chinese economy, and raises concerns about the ability of the People's Bank of China (PBoC) to control the Renminbi (RMB) and about the macroeconomic consequences of the reform.


A. The Dual Exchange Rate System

China's currency, the RMB, has been inconvertible since it became a national currency in 1949. Until the start of the economic reform in 1978, the central government directly planned China's foreign trade and foreign exchange allocation. A dozen of foreign trade corporations (FTCs) carried out foreign trade plans. The FTCs had to surrender all their foreign exchange earnings to and purchase foreign exchange from the central government at the official exchange rate. Besides receiving FTC profits, the central government subsidized FTC losses. During this period, the RMB exchange rate served little economic function. It was only a price for budget allocation under the trade plans.

The country's foreign trade reform has driven the development of the exchange rate system since 1978. In order to increase the role of market forces and to reduce the burden of trade subsidies on the central government's budget, China started in 1979 to decentralize foreign trade and the administration of foreign exchange earnings. It also gradually reduced the fiscal subsidies for the FTCs and introduced a foreign exchange retention system. Under this system, FTCs gained the right to buy back at the official exchange rate specified quotas of foreign exchange earnings previously surrendered to the central government. The retention system led to establishing the FEACs across the country in 1986. Table 1 summarizes the FEACs' main features. Through the FEACs, enterprises short of foreign exchange could "adjust" their needs by purchasing at a negotiated rate foreign exchange quotas from the enterprises who had surplus foreign exchange quotas. China thus had a dual exchange rate system. It used the administered official exchange rate for transactions under the foreign trade plan and for most capital account transactions. On the other hand, the more depreciated, flexible swap market exchange rates determined in the FEACs were applicable to foreign exchange transactions outside the foreign trade plan.

B. The Unified Foreign Exchange System

On January 1, 1994, China unified its official and swap market exchange rates (see Lo and Bang, 1994; PBoC, 1994a, 1994b; Zhou, 1994). Table 2 summarizes the details of the reform process. At end of 1993, the official exchange rate was depreciated from RMB 5.8 to RMB 8.7 per U.S. dollar, the Shanghai swap rate at the time. (All exchange rates below are quoted as RMB/USD.) In April 1994, the China Foreign Exchange Trading System (CFETS) was established in Shanghai. A national interbank foreign exchange market centered around the designated foreign exchange banks (DFEBs) replaced the system of local and unlinked FEACs. By the end of 1994, the CEFTS had 303 members from 26 cities and produced a cumulated transaction volume of USD 40.7 billion from April to December 1994. The new regime abolished the foreign exchange retention system and required domestic enterprises to sell all their foreign exchange earnings to and purchase foreign exchange from the DFEBs at the exchange rates quoted by the banks. These exchange rates were quoted around the "middle" rate set by the PBoC according to the previous trading day's supply and demand conditions. The DFEBs are allowed to keep a foreign exchange balance to meet customer demand. However, this balance must fall within a certain percentage of the DFEBs' foreign exchange assets. The DFEBs must eliminate any surplus or shortfall by trading with other DFEBs through the interbank market or with the PBoC (see Asiamoney, 1994). The new system also enhances the RMB's current account convertibility by abolishing the authorization process in acquiring foreign exchange for normal current account transactions.


A. Sources of Market Segmentation and Linkages

The FEACs were local organizations that operate independently from one another. They were subject to regulations set by local branches of the State Foreign Exchange Administration Bureau (SFEAB), which is subordinated to PBoC. The SFEAB generally did not buy and sell foreign exchange in the FEACs except when a significant gap separated the swap and official exchange rates. The SFEAB exerted its control over the swap rates though administrative measures to restrict access to the FEACs. Generally speaking, sales of foreign exchange were subject to few restrictions. In contrast, stringent restrictions applied to the purchasing side. Upon presentation of valid documents, such as import licenses, the SFEAB would check the foreign exchange uses for conformity with the priority list set by the government before approving or rejecting foreign exchange purchases.

The fact that the FEACs limited membership to local participants created price differentials among the different FEACs' swap rates. Several other swap market features also contributed to market segmentation. First, local governments frequently intervened to prevent foreign exchange in their jurisdiction from flowing to other regions, since foreign exchange was an underpriced and scarce resource. This problem was especially serious during periods of tight foreign exchange supply. Wang (1992) and Tang (1993) document evidence of this type of administrative control for the Anhui and Liaoning Provinces.

Second, systematic information flows among FEACs were nonexistent since the FEACs were not linked together by any communication network. In most cases, prevailing prices in other FEACs were not available to buyers. According to Zhiping Liu (1991), "among the many swap centers, only a few of them will periodically inform each other of their prices. The centers only report their trading prices and volume to the Exchange Rate Division of the SFEAB." The SFEAB, however, was an administrative authority and did not publish the information in a timely manner.

Third, the FEAC administrative structure also created segmentation. The prefectural FEACs were subordinated to the provincial FEACs of their own provinces. Transactions between prefectural centers of different provinces had to be conducted through their respective provincial FEACs. This administrative procedure discouraged inter-regional transactions and impeded the efficiency of foreign exchange allocation.

Despite the segmentation, some linkages - direct or indirect - did exist. The liberalization of the foreign trade sector made an increasing number of FTCs responsible for their own domestic currency profits and losses. This provided economic incentives for enterprises and local governments to compete for foreign exchange from FEACs across the country. For example, in 1990 the supply could not satisfy demand in Wuxi city, and the Wuxi FEAC was able to obtain two million dollars from nearby cities (Huang, 1990). In 1993, the XinJiang FEAC adjusted almost one million dollars from other provinces when an acute shortage of foreign exchange occurred in the XinJiang Province (Financial Times, 1994).

Cross-regional investment and operation of FTCs and foreign investment enterprises (FIE) also provided a channel for linkage among FEACs. Despite the effort to decentralize the foreign trade sector, trade enterprises with cross-regional connections remained the dominant players in the swap market. In regions with large-scale foreign investment, FIEs also were very important. For example, of the 43 dealers in the Shanghai FEAC in 1990, 35 were state and collective businesses and nine were FIEs (Shanghai Swap Center, 1991). In the Dalian FEAC, FIEs alone produced 50% of the supply of foreign exchange (Financial Times, 1993). Since foreign exchange bought from the FEACs must be used according to the purposes specified in the purchasing application, policy prohibited direct arbitrage to buy low in one center and sell high in another center. However, an enterprise with operations in different regions could arbitrage by selling foreign exchange at FEACs that offered a higher rate and buying foreign exchange at a lower rate in other regions for imports. The movement of goods provided another way to arbitrage. A multi-regional enterprise could import from a region where the swap market exchange rate was low and ship the imports to a destination where cost of foreign exchange was high.

B. Degree of Market Segmentation

This subsection assesses the degree of the swap market segmentation using the biweekly swap rates data of the 12 most active FEACs in China from the beginning of 1992 to the end of 1993. Except where specified otherwise, data are from Financial Times (China), various issues; China Finance, various issues; International Financial Statistics, various issues; Statistical Yearbook of China, various issues; China Financial Outlook, People's Bank of China (1994); China: Foreign Trade Reform and China: Internal Market Development and Regulation, World Bank (1994). The 12 centers are the National Center (NC), Beijing (BJ), Tienjin (TJ), and Dalian (DL) in the north; Shanghai (SH), Nanjing (NJ), and Hangzhou (HZ) in the east; Guangzhou (GZ), Shenzhen (SZ), and Xiamen (XM) in the south; and Xian (XI) and Chengdu (CD) in the west.

First, examine the degree of the swap market's segmentation, which is measured by the absolute exchange rate differentials between the FEACs' swap rates. Table 3 presents the 66 pairwise mean absolute differentials (MADs) among the 12 FEACs. These MADs, expressed as a percentage of the two FEACs' average swap rates, range from 1.2% to 3.5%, with a grand mean of around 2%. All the MADs are statistically significant at the 1% level. Furthermore, these exchange rate gaps between FEACs do not appear to narrow over time. Regressions of these absolute exchange rate differential series on a constant and a time trend indicate that all the time trend coefficients are very small and statistically insignificant. This suggests that important administrative and information barriers remained to inhibit the swap market's development and integration. Closely examining the figures in table 3 suggests that the degree of market segmentation varies. For example, the MADs between the northern and eastern FEACs are generally smaller than those between the northern and southern FEACs. The null hypothesis that all the 66 pairwise absolute differential series have the same MADs is rejected in the ANOVA test at the 1% level, with [F.sub.65,3960]=2.14.

The diversity in the degree of market segmentation becomes more apparent at the regional level. Panel I of table 4 summarizes the MAD figures in table 3 at the regional level. Note that the regional MADs are measured in levels here, not in percentage terms. Several observations emerge. First, all regional MADs are statistically significant at the 1% level. Hence, the swap market also was segmented at the regional level. Second, in terms of own region MADs (i.e., the four diagonal cells), the eastern region is the most integrated among the four regions while the southern region is the least integrated. Of the six inter-regional MADs, the northern-eastern regional MAD is the smallest, suggesting that FEACs in these two regions are more integrated than are FEACS in the remaining regions. On the other hand, the eastern-southern regional MAD is the largest. Finally, the southern region appears to be isolated from the other three regions. The southern region exhibits the largest mean absolute differential from each of these three regions.

Confirming the statistical significance of the observed regional patterns involves several tests. First, the null hypothesis that all the 10 pairwise regional absolute differential series have the same MADs is rejected in the ANOVA test at the 1% level, with [F.sub.9,4016]=5.88. Next, the analysis tests [TABULAR DATA FOR TABLE 3 OMITTED] the null hypothesis that pairwise regional MADs in panel I of table 4 are equal. Since there are 10 regional MADs, the analysis conducts 45 t-tests. Panel II of table 4 presents the results of the six tests that involve only two own-region MADs. The western region is the only region where one cannot reject the null hypothesis that its own-region MAD is the same as the other three regions' own-region MADs. As for the other three regions, their own-region MADs differ significantly from one another's. Of the remaining 39 pairs (not reported in table 4), the null hypothesis is rejected in 26 of them.

Finally, the analysis compares relative levels of the 12 FEACs' swap rates. For this purpose, each period exhibits the sign ([greater than], [less than], or =) of the difference between the individual center's swap rate and the 12 FEACs' average swap rate in that period. Panel I of table 5 summarizes these results. Note that the eastern and western regions' swap rates are greater than the average more often than they are less than the average rate. On the other hand, the southern FEACs' exchange rates are below the average rate more often. Panel II of table 5 presents each of the 12 FEACs' mean exchange rate. Again, the FEACs' swap rates in the eastern and western regions are higher than those in the southern region, with the swap rates of the northern region lying in between.

These results show evidence of market segmentation in China's swap market. Informational and administrative barriers apparently continued to cause substantial differences in the FEACs' swap rates across the country. The contrasts between the eastern and southern regions are especially sharp.

C. Linkage and Causal Relationship Between FEACS

This subsection examines the causal relationship between the 12 FEACs' swap rates and the three reference centers - namely Beijing, Shanghai, and Shenzhen. First, the analysis tests for the swap market exchange rates' order of integration using the Augmented Dickey Fuller test. The test indicates that the series are integrated of order one since all the null hypotheses of a unit root in the series cannot be rejected in the level but are rejected in the first-difference. This finding suggests that swap rates may be pairwise cointegrated. The two-step approach proposed by Engle and Granger (1987) reveals that, except for the swap rates in the GZ-XI and the GZ-SH pairs, the remaining swap rate pairs are cointegrated. Cointegration implies existence of an error correction mechanism (ECM). Therefore, the analysis tests for a causal relationship in the first difference for the two non-cointegrated pairs whereas for the other cointegrated pairs it tests for causality relationship using an ECM.

Suppose the swap rates from a reference center X and an individual center Y are cointegrated. The ECM will take the following forms:

(1) [Delta][X.sub.t] = [[Alpha].sub.1] + [[Rho].sub.1][u.sub.t-1]

+ [summation of] [[Gamma].sub.i][Delta][X.sub.t-i] where i=0 to n + [summation of] [[Delta].sub.j][Delta][Y.sub.t-j] where j=0 to n + [[Epsilon].sub.1,t]

(2) [Delta][Y.sub.t] = [[Alpha].sub.2] + [[Rho].sub.2][v.sub.t-1]

+ [summation of] [[Beta].sub.i][Delta][X.sub.t-i] where i=0 to m + [summation of] [[Gamma].sub.j][Delta][Y.sub.t-j] where j=0 to m + [[Epsilon].sub.2,t]

where [a.sub.1], [a.sub.2] are the intercept terms, [u.sub.t-1] and [v.sub.t-1] are the error-correction terms, and [[Epsilon].sub.1] and [[Epsilon].sub.2] are the random errors. [u.sub.t-1] is the residual of the cointegrating regression with X as the dependent variable and Y as the independent variable. [v.sub.t-1] is the residual from the reverse cointegrating regression.

Note that one variable can affect the other through one of two channels. In terms of equation (1), the first channel is the lag difference channel from the [Delta][Y.sub.t-j] terms. The statistical significance of the [[Delta].sub.j] coefficients implies Granger causality from Y to X. The second channel is the error-correction channel from the error-correction term [u.sub.t]. The coefficient [[Rho].sub.1] measures the adjustment of dependent variable X to short-run departures from the long-run equilibrium relationship embodied in the cointegrating regression [X.sub.t] = a + [Beta][Y.sub.t] + [u.sub.t]. The statistical significance of [[Rho].sub.1] implies that one-period-lagged value of Y can help forecast the current value of X. In other words, Y Granger-causes X. Table 6 and Table 7 present the results of these two causal relationship channels. Table 6 reports the coefficient estimates of [[Rho].sub.1] and [[Rho].sub.2] while Table 7 reports the results of the F test for the null hypothesis that [[Delta].sub.j]s in equation (1) or [[Beta].sub.i]s in equation (2) are equal to zero.

Mean Absolute Differentials Between Regions

I. Regional Mean Absolute Differentials(1)

 North East South West

North .132(***)
East .129(***) 0.94(***)
 (19.7) (6.93)
South .158(***) .184(***) .172(***)
 (19.5) (16.8) (10.1)
West .145(***) .134(***) .175(***)
 (19.9) (14.1) (14.4) (6.7)

II. t-ratio of the differentials between own-region MADS(2)


EE 2.46(***)
SS 2.38(***) 3.56(***)
WW 0.05 -1.43 1.27

1 The upper figure of each cell in panel I is the MAD between the
column region X and row region

 [Mathematical Expression Omitted]

where n = number of FEACs in region X

m = number of FEACs in region Y

[MAD.sub.ij] = mean absolute differential between swap center i in
region X and swap center j in region Y

The lower bracketed figure of each cell in panel I is the t
statistics of the MAD between the column and row regions.

2 The t-ratio of each cell tests the null hypothesis that the MAD
within the FEACs in the column region is equal to that of the row
region, e.g., the t-ratio in the NN - EE cell (2.46) tests if the
MAD in column 1 and row 1 of panel I (0.132) is equal to the MAD in
column 2 and row 2 of panel I (0.094).

*** denotes significance at 1% level.

Several aspects of these results are worth noting. First, contrary to intuition, in quite a few cases Shanghai reference center's swap rates are Granger-caused by the swap rates of smaller and less active individual center's rates. This pattern also exists, although to a less extent, in the case of the Beijing reference center. This counter-intuitive result probably reflects that foreign exchange supplies in smaller FEACs usually were tighter than those in larger, more well-developed FEACs. As a result, a change in a smaller center's swap rate could induce a change in a larger center's swap rate when participants in the smaller center tried to purchase foreign exchange from the larger centers. The results suggest that many other FEACs attempted to adjust their foreign exchange needs through the Shanghai FEAC market. Note that the main channel of causality from other centers to Shanghai is the ECM channel, which implies that the Shanghai swap rates adjusted to short-run deviations from long-run equilibrium relationship. By comparison, the Beijing reference center exhibits a feedback relationship with FEACs from its own region as well as from the eastern region and also is Granger-caused by FEACs of other centers through the lag difference channel and the ECM channel. Finally, Shenzhen FEACs' swap rates Granger-cause other FEACs' swap rates through both channels. The existence of these causality relationships indicates linkages between FEACs. Nevertheless, such linkages were not strong enough to eliminate or even to narrow the gaps between the regions. They only served to eliminate short-run deviations from the long-run relationships.

Comparison of the Exchange Rate Levels of the 12 FEACs

I. Frequency of Signs of Deviations(1)

Region Center National Average

 [greater than] = [less

North NC 23 0 38
 BJ 22 1 38
 TJ 40 0 21
 DL 23 1 37

East SH 42 0 19
 NJ 47 0 14
 HZ 45 0 16

South GZ 22 0 39
 SZ 21 0 40
 XM 13 0 48

West XI 43 0 18
 CD 35 0 26

1 For each individual center, the three numbers reported are, in
order, the frequency that its swap is greater than ([greater
equal to (=), or less than ([less than]) the natural average over
the sample period (61 observations),

II. Mean exchange rate of individual FEACs

Region Center Mean

North NC 7.66
 BJ 7.68
 TJ 7.74
 DL 7.69

East SH 7.75
 NJ 7.80
 HZ 7.74

South GZ 7.64
 SZ 7.67
 XM 7.55

West XI 7.80
 CD 7.70




A. Implications at the Micro Level

The new exchange rate system has greatly improved foreign exchange allocation efficiency in several ways. First, the nationally unified market rate has eliminated the persistent differentials among different FEACs' swap rates. Second, participants in the unified foreign exchange market benefit from the freedom to choose any DFEB to trade with and no longer are restricted to the local FEACs in their areas. Third, the geographical coverage and size of the interbank network free the unified exchange rate from price manipulation by a few big players. Fourth, the coverage of the market exchange rate has expanded to include all the current account transactions.

The new system has enhanced the ability of PBoC to control the exchange rate by channelling all the foreign exchange to the DFEBs. Elimination of the retention system has stripped the FTCs of their retained quotas, which could be used to manipulate the market. Furthermore, under the new regime, the SFEAB must approve and supervise the opening of foreign exchange accounts. Deposits into and payment from these accounts must be in line with the prescribed scope that the SFEAB has laid down. Most domestic enterprises' foreign exchange accounts gradually are being abolished, and billions of U.S. dollars previously held in these accounts now are concentrated in the DFEBs.

The unified interbank network has increased the effectiveness and reduced the administrative costs of central bank market intervention. The new system has given the PBoC a greater degree of freedom in controlling the exchange rates. For instance, the central bank can control the "middle rate" around which DFEBs quote their rates as well as the margin around that rate. Alternatively, the degree of exchange rate flexibility also can be adjusted by varying the foreign exchange exposure ceilings of the DFEBs. In the worst case, quantity still can be restricted since the DFEBs are under the direct supervision and inspection of the SFEAB.

B. Implications for the Macro Economy

The macroeconomic consequences of the unification in China are quite different from those of other countries such as Zambia, Uganda, and Sierra Leone that undertook foreign exchange market unification in recent years (Roberts, 1989). That is, in China, more expensive imported goods and/or an increase in budget deficit were not the major sources of inflationary pressure after the unification (Pinto, 1991). On the surface, the 50% depreciation (from 5.8 to 8.7) in the official rate on January 1, 1994, could have led to a drastic increase in inflation. In fact, the inflationary impact was much less significant than the depreciation would suggest. Before the unification, the average foreign exchange retention rate was about 80%. The retained foreign exchange could be used for importing or for sales in the swap markets. Therefore, only about 20% of the current account related foreign exchange transactions were made under the official rate and hence could have been affected by the 50% depreciation of the official rate.

Effects of the official rate depreciation on the central government budget were small. Before 1992, the government heavily subsidized foreign trade corporations' losses by covering losses of both exports and imports. For example, the subsidies were 34 billion Yuan in 1989, 22 billion Yuan in 1990, and 18 billion Yuan in 1991 (World Bank, 1994). However, starting from 1992, the government suspended such subsidies totally. As a compensation, the government significantly raised the foreign exchange retention rate from an average of under 50% to 80%. Trading the retained foreign exchange at the much depreciated swap rates provided a relief to the exporting firms. Meanwhile, the importing firms no longer were entitled to cheap foreign exchange or government subsidies.

The unification's main impact on the money supply came from the central bank's unsterilized foreign exchange purchase. The new system abolished foreign exchange retention and required exporting firms to sell all foreign exchange proceeds to the DFEBs. This regulation increased the foreign exchange supply in the market. At the same time, a tightening of bank credit to enterprises in 1994 contributed to the excess supply of foreign exchange and excess demand for RMB in the market. The confidence in the Chinese economy and the significant interest rate spread between RMB and USD interest rates also brought about a surge in capital inflows. Since the DFEBs were allowed to hold only a limited amount of foreign exchange, they were forced to sell the extra foreign exchange to the PBoC. China's international reserves rose from USD 20 billion at the end of 1993 to over USD 50 billion at the end of 1994. The capital inflows induced by the unification partially reduced the PBoC's ability to control the base money and the inflation target. Therefore, monetary policy has become more influenced by external conditions.


The new foreign exchange system has enhanced the current account convertibility through abolishing the approval procedure for acquiring foreign exchange for normal current account transactions (Tsang, 1995). The old system required importers to obtain both trade approval and authorization to purchase foreign exchange from their local FEACs. The new system allows importers to purchase foreign exchange from any DFEB once they obtain trade approval. However, the progress towards convertibility actually is more superficial than real since DFEBs are under direct supervision of the SFEAB, which has not relaxed restrictions on the uses of foreign exchange that prevailed before the unification. On the supply side of foreign exchange, the degree of convertibility actually has decreased since the retention quota system was abolished.

Establishing currency convertibility is rarely a one step matter; it often occurs in stages. Removing restrictions on current account transactions could pose serious risks for macroeconomic stability, particularly early in the reform process when the preconditions for convertibility have not been in place long. These preconditions (Greene and Isard, 1991) include (i) an appropriate exchange rate level that is consistent with balance of payments equilibrium, (ii) an adequate level of foreign exchange reserves, (iii) effective fiscal and monetary policy instruments to ensure macroeconomic stability, and (iv) a market environment where producers and consumers have the incentive and ability to respond to market signals. Of the four pre-conditions, China probably could satisfy only the first two. However, China recently has adopted policies to fulfil the last two. Abolishing all subsidies to FTC in 1992 represented substantial progress in foreign trade reform, but China's state enterprise reform has just started. Most state enterprises still operate under soft budget constraints. Their losses have imposed a constraint on monetary policy and put pressure on the budget deficit. Hence, China is not yet ready for full current account convertibility.

The 1994 foreign exchange reform, however, has contributed to establishing the preconditions for convertibility. First, somewhat ironically, the reform has led to a major increase in foreign exchange reserves. The new exchange rate regime has channelled all foreign exchange to the interbank market and enabled the central bank to control the exchange rate more effectively and at a lower administrative cost. Under the FEACs system, most tradings were in the form of retention quotas instead of cash. The PBoC was not able to intervene effectively in the foreign exchange market in order to stabilize the sometimes volatile exchange rate. The new system has solved the problem of FTCs' hoarding foreign exchange retention quotas for speculative purposes. Thus, the new exchange rate system has strengthened the central banks' ability to stabilize the exchange rate and the current account and to mitigate some of the risks associated with currency convertibility. The unified system provides a more efficient trading and settlement mechanism and a unified regulatory framework. This is an effective step toward making the Renminbi fully convertible.


The exchange rate unification experience in China is rather unique compared with some developing countries. For example, in Sierra Leone, Guinea, Somalia, Ghana, and Zambia, unification generally led to drastic increases in the budget deficit, and in the domestic money supply and finally to a burst in inflation. Broad money and the price level in these countries grew by 40% to 60% in the first year following the unification. In contrast, China's 1994 M2 growth rate was 33% and her inflation rate was 22%, which are relatively mild compared to those of other developing countries. Remarkably, China's exchange rate has been quite stable after the unification. The policy of gradually liberalizing the foreign exchange market by establishing a "semi-floating" swap market and by increasing its coverage overtime seems to have protected China from the budget deficit 'and hyper-inflation problems that usually result from exchange rate unification.

Although the institutional structure of a bank-based foreign exchange market has been established, it remains to be seen whether the reform is a real success. If the PBoC heavily manages the so-called market rate using administrative instruments, progress towards full convertibility will suffer a serious setback. This will create a situation worse than the swap market era, because no swap market will be available to adjust the excess demand for foreign exchange. A black market for foreign exchange - which is very costly to control - will re-emerge. The credibility of the reform will be damaged, increasing the risk of capital flight. The success of the unification ultimately depends on how completely and how quickly the government removes restrictions on market access and on how well decision makers coordinate the monetary and fiscal policies with the move towards higher convertibility.


CFETS = China Foreign Exchange Trading System

DFEB = Designated foreign exchange bank

ECM = Error-correction model

FEAC = Foreign exchange adjustment center

FIE = Foreign investment enterprise

FTC = Foreign trade corporation

MAD = Mean absolute differential

PBoC = People's Bank of China

RMB = Renminbi

SFEAB = State Foreign Exchange Administration Bureau

USD = U.S. dollar

BJ = Beijing

CD = Chengdu

DL = Dalian

GZ = Guangzhou

HZ = Hanzhou

NC = National Center

NJ = Nanjing

SH = Shanghai

SZ = Shenzhen

TJ = Tienjin

XI = Xian

XM = Xiamen


Main Features of China's Foreign Exchange Swap Market


All state, collective, foreign-investment, and private enterprises and businesses in the region may participate. Self-employed business people and households may sell their foreign exchange in this market, but they are not allowed to buy foreign exchange. The central bank is represented in the market. It buys and sells foreign exchange to stabilize demand and supply around certain price range.


A membership system is adopted. Only members may bid and ask in the market. Non-members cannot trade directly in the centers, and must trade through their brokers. There are two types of memberships:

(1) Brokers. They are banks and non-bank financial institutions that hold a Foreign Exchange Businesses Permit issued by the SFEAB. They can only trade as agents for their clients.

(2) Dealers. These are non-financial institutions that participate in foreign exchange operations. They can only trade for themselves.

Applications for membership must be submitted in writing to the local branch of the SFEAB. Upon approval, each member possesses one seat in the center and may send up to two representatives, who must be holders of a certificate also issued by the SFEAB.

Qualified sources of foreign exchange

Retained foreign exchange quotas and foreign exchange by foreign exchange earning enterprises engaged in trade and non-trade activities; foreign exchange income of the foreign investment enterprises; and foreign exchange owned by self-employed business people and individuals.

Qualified uses of foreign exchange

Top priorities: the import of agricultural input goods, necessary goods for living, input goods for key projects, foreign exchange earning production, high-tech equipment, repayment of principal and interest of foreign debts, and profit repatriation.

Secondary priorities: input goods for industrial production, goods for research, education, health care, culture, business needs of foreign investment enterprises, repayment of foreign currency loans made by domestic financial institutions, etc.

Typically, import of luxury consumption goods and funding for foreign exchange investment of domestic enterprises in and outside the country are not entertained.

Currencies traded

U.S. dollar, Great British pound, German mark, French franc, Japanese yen, and Hong Kong dollar. Foreign exchange quotas are denoted in the U.S. dollar.


The price is determined principally by demand and supply conditions through a bidding process. The price for quotas is determined as the difference between the U.S. dollar (cash) swap rate and the current official price.

Trading process

Both the buyers and sellers must get approval in advance from the local branch of the SFEAB and must hold a certificate certifying a specific sales or purchase. The buyers and sellers can then either apply for trading directly or ask a broker to make a trading application to the swap center, upon presenting the SFEAB certificate.

The swap center will pool all the qualified. applications together and have trading sessions. The member's name, trading quantity and price will be published. When the bid price is lower than the ask price, the bidding will go on, and the bidder and asker will adjust their original prices. When the bid price is higher than the ask price, the highest bid wins.

Delivery and settlement

A sum of guarantee money must be paid up before the trading. The broker is responsible for non-member's delivery and settlement. The swap center takes care of dealers' delivery and settlement. So the non-members and the dealers should deposit enough RMB or foreign exchange guarantee money into special accounts held with the broker and the center, respectively.

Settlement of transactions must be completed before noon of the next business day.


Main Features of the Unification of Exchange Rates in 1994

Main contents of the reform package

(1) Unifying the exchange rates, and implementing a market-based, managed-floating exchange rate. The official rate was depreciated from RMB 5.8 to US$1 at end of 1993 to RMB 8.7 to US$1 (which was the swap rate at that time) on January 1, 1994. The official rate has been determined principally by the foreign exchange market ever since then.

(2) Replacing the foreign exchange retention system with a system whereforeign exchange should be sold to and bought from the commercial banks (under the retention system, foreign exchange was sold to and bought from the central bank). Compulsory plans governing the foreign exchange income and uses were abolished. The demand for and supply of foreign exchange were matched through the foreign exchange market instead of through the plans.

(3) Establishing a nationally unified inter-bank foreign exchange market to replace the local and segmented swap centers, and introducing a computerized trading and information system. This market helps adjust the foreign exchange positions of the banks and provides clearing services.

The first stage: January 1 to March 31, 1994

Before a well-established foreign exchange market that is centered on banks was established, the major swap centers across the country were retained for transitory purposes. In the period January 1, 1994 to March 31, 1994, the exchange rate was determined as follows. The swap centers remained functioning, but they had to refer to the rate of the swap market in Shanghai in determining their own rates. The regional centers should be active in trading with other regions. This regulation largely unified the earlier different rates across the country, leaving only minor variations. The official RMB to U.S. dollar rate was published by the People's Bank everyday, the calculation of which was based on the weighted average price of 18 swap centers throughout the country in the earlier business day. All foreign exchange transactions were settled according to this price.

The second stage: April 1 to July 31, 1994

(1) The market is centered around banks instead of around the swap centers. Foreign exchange should be sold to the designated banks and be bought from them (providing they hold valid certificates). The banks form an inter-bank foreign exchange market in adjusting their positions. The central bank intervenes in the inter-bank market by buying and selling to affect the exchange rate. The swap centers will phase out gradually.

(2) The official exchange rate is determined daily based on the weighted average inter-bank rate. The designated foreign exchange banks buy and sell foreign exchange within a floating range of the official rate. The inter-bank market guarantees that rates of different banks will not differ by a large margin.

(3) Computerized national trading system was established to connect different regions and banks, and to speed up clearing and settlement.

(4) National-unified regulations on the foreign exchange market were put into operation.

The third stage: post July 1994

Foreign investment enterprises were adopted into the foreign exchange market centered on banks. In the first and second stages, they remained buying and selling foreign exchange through the swap markets and were allowed to continue to keep their foreign exchange in designated accounts (instead of selling it to a bank). This was meant to guarantee the access of foreign investment enterprises to foreign exchange and to consolidate their confidence before the new system was well-developed.


Asianmoney, Asian Currency Guide, Supplement, 1994, 8-11.

Engle, R. F., and C. W. Granger, "Cointegration and Error Correction: Representation, Estimation, and Testing," Econometrica, 55:2, 1987, 251-276.

Greene, J., and Peter Isard, "Currency Convertibility and the Transformation of Centrally Planned Economies," International Monetary Fund Occasional Paper, no 81, 1991.

Financial Times (China), December 3, 1993.

-----, May 20, 1994.

Huang, Wanlian, "An Analysis of Moves of the Foreign Exchange Swap Market in the Post-Depreciation Period," China Finance, 6, 1990, 46-47.

Liu, Zhiping, "The Main Problems and Their Solutions of China's Foreign Exchange Swap Market," China Finance, 1, 1991, 45-47.

Lo, Elaine, and E. Bang, "Provisions Bring in New Forex Regime for FIES," China Banking and Finance, 8, 1994, 8-10.

People's Bank of China, Further Reform of the Foreign Exchange Control System Announcement, 1993.

-----, China Financial Outlook, 1994(a).

-----, "Administration of Foreign Exchange Provisions Full Text," China Banking and Finance, 23, May 1994(b), 4-9.

Pinto, Brian, "Black Markets for Foreign Exchange, Real Exchange Rates and Inflation," Journal of International Economics, 30:1, 1991, 121-135.

Roberts, John, "Liberalizing Foreign-Exchange Rates in Sub-Saharan Africa," Development Policy Review, 7, 1989, 115-142.

Shanghai Swap Center, "Exploring and Improving Gradually the Foreign Exchange Swap Market," China Finance, 7, 1991, 32-33.

Tang, Pinghuang, "Further Improving and Developing the Foreign Exchange Market," China Finance, 7, 1993, 30-31.

Tsang, Shu Ki, "The Banking and Foreign Exchange Reforms in China: Review and Assessment," Business Research Center Paper on China, no. CP95006, 1995, Baptist University of Hong Kong.

Wang, Baoquan, "Improving the Operational Mechanism of the Foreign Exchange Swap Market," China Finance, 6, 1992, 45.

Zhou, Yingyo, "A Fundamental Reform of the Foreign Exchange Administration System," China Finance, 3, 1994, 17-19.

World Bank, China: Foreign Trade Reform, The World Bank, Washington, D.C., 1994.


Assistant Professors, Department of Economics and Finance, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, Hong Kong. This is a revised version of a paper presented at the 69th Western Economic Association International Annual Conference, Vancouver, B.C., July 1, 1994, in a session organized by Gene Hsin Chang, University of Toledo, Ohio, and Jack W. Hou, California State University, Long Beach. A research grant from the City University of Hong Kong supported this study. The authors thank Professor Lilian Ng, participants of the WEA International sessions, and three anonymous referees for helpful comments on earlier drafts.
COPYRIGHT 1996 Western Economic Association International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Huang, Guobo; Wong, Yuk-Pang
Publication:Contemporary Economic Policy
Date:Oct 1, 1996
Previous Article:Politics and length of time to bank failure: 1986-1990.
Next Article:Competitive need limits and the U.S. generalized system of preference.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters